The FIND project
Foundation for Industry (FIND) – Large AI Models for a Resilient High-Tech Industry – is a 6-year nationally funded programme uniting 5 Dutch universities, 2 knowledge institutes, and 11 industrial partners. Its aim is to develop and apply foundation models tailored to Dutch industry needs, focusing on underserved modalities (e.g. sensor data, time series, multimodal signals) and edge applications. The consortium develops an open-source toolbox of models, code, and validated outputs to strengthen the competitiveness of the Dutch high-tech sector.
The PhD Candidate Position
You will be supervised by Prof. Dr. Suzan Verberne and Dr. Niki van Stein at the Leiden Institute of Advanced Computer Science (LIACS), Leiden University. The focus is on explainability for multimodal foundation models, combining information-theoretic approaches (e.g., information bottleneck) with cross-modal modelling of time series and sensor data. The project will also explore signal-to-text generation for human-readable explanations of complex sensor outputs and mechanistic interpretability for multimodal foundation models.
Key responsibilities
- Perform PhD research on the above topics and publish in top venues;
- Collaborate closely with the other FIND fellows and industrial partners;
- Disseminate results via publications, open-source software, and presentations;
- Assist in relevant teaching activities.
Selection Criteria
- MSc degree in Computer Science, Artificial Intelligence, Data Science, or related field;
- Solid background in machine learning, deep learning and foundation models such as Large Language Models;
- Strong programming skills (Python/C++);
- Proven interest in generative models and Explainable AI is a plus;
- Analytical and creative mindset, with the ability to work independently and in teams;
- Excellent command of English; willingness to work in an international, multi-sector environment;
- Good collaboration skills, also in interdisciplinary settings.